Fault Detection of Gear Box using Artificial Neural Network (ANN)

Fault diagnosis plays an important role in condition monitoring to enhance the machine time. In view of this, the present investigation focused on the development of Fault diagnosis system of gear boxes based on the vibration signatures and Artificial Neural Networks. In the present investigation to generate the vibration signatures an experimental set-up has been fabricated with sensing and measuring equipment. The four prominent faults wear,crack, broken tooth and insufficient lubrication of the gear were practically induced in the present investigation. Vibration signatures of the gearbox were collected by transmitting the motion at constant speed with gears having no fault, without applying any load.By inducing one fault at a time, vibration signatures were collected with different faults. Further, it was decided to use ANN based fault diagnosis system for the present investigation. The set of statistical features were extracted based on data pertaining to maximum amplitudes of vibration. The lower frequency of vibration Signal (RMS) is used as input to the ANN based fault diagnosis system designed and developed in MATLAB.